# GPT-6 Luna vs Grok 4.20 Multi-Agent

> GPT-6 Luna is the stronger model overall, scoring 53.3 to 46.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-6-luna-vs-grok-4-20-multi-agent
- Last updated: 2026-10-12
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. GPT-6 Luna scores higher in 5 categories and Grok 4.20 Multi-Agent in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Luna leads 76.1 to 39.4.
- The biggest single-benchmark swing is NYT Connections (extended): 68.7% for GPT-6 Luna and 89.6% for Grok 4.20 Multi-Agent.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $1.25 / $2.50 for Grok 4.20 Multi-Agent.
- GPT-6 Luna accepts more context: 1.05M tokens versus 1M.

## Snapshot

| | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 53.3 | 46.2 |
| Rank | 36 | 65 |
| Context | 1.05M | 1M |
| Input $/M | $0.10 | $1.25 |
| Output $/M | $0.50 | $2.50 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-6 Luna: 55.5 (#25)
- Grok 4.20 Multi-Agent: 43.0 (#92)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Coding | 1439 | 1457 |
| DeepSWE | 66.6% | — |
| FrontierCode | 42.4% | — |
| LMArena WebDev | 1581 | — |
| SciCode | 54.6% | — |
| ALE-Bench | 1,577 | — |

## Agentic & Tool Use

- GPT-6 Luna: 33.3 (#54)
- Grok 4.20 Multi-Agent: —

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| APEX-Agents | 44.3% | — |
| GDP.pdf | 23% | — |
| LMArena Search | — | 1204 |

## Reasoning

- GPT-6 Luna: 48.2 (#41)
- Grok 4.20 Multi-Agent: 43.9 (#48)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| NYT Connections (extended) | 68.7% | 89.6% |
| LMArena Hard Prompts | 1411 | 1448 |
| ARC-AGI-2 | 59.3% | — |
| ARC-AGI-1 | 86.7% | — |
| CritPt | 19.4% | — |
| Chess Puzzles | 31% | — |
| Mystery Game Puzzles | 7% | — |
| DTBench | 90.1% | — |
| LMCA | 44.5% | — |
| Epoch Capabilities Index | 156.28 | — |

## Math

- GPT-6 Luna: 76.1 (#15)
- Grok 4.20 Multi-Agent: 39.4 (#104)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Math | 1416 | 1442 |
| FrontierMath (Tiers 1-3) | 78.9% | — |
| FrontierMath Tier 4 | 56.1% | — |
| OTIS Mock AIME 2024-2025 | 98.9% | — |
| ProofBench | 64% | — |

## Knowledge

- GPT-6 Luna: 57.0 (#41)
- Grok 4.20 Multi-Agent: 40.4 (#119)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Expert | 1444 | 1445 |
| GPQA Diamond | 90.5% | — |
| SimpleQA Verified | 41.4% | — |

## Multimodal

- GPT-6 Luna: 42.4 (#30)
- Grok 4.20 Multi-Agent: 40.5 (#48)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Vision | 1217 | 1259 |
| Blueprint-Bench 2 | 31.2% | — |
| Furniture Assembly | 44.2% | — |

## Multilingual

- GPT-6 Luna: 50.5 (#117)
- Grok 4.20 Multi-Agent: 54.4 (#43)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Non-English | 1386 | 1440 |
| LMArena Chinese | 1433 | 1475 |
| LMArena French | 1420 | 1466 |
| LMArena German | 1369 | 1456 |
| LMArena Japanese | 1369 | 1405 |
| LMArena Korean | 1360 | 1416 |
| LMArena Russian | 1394 | 1457 |
| LMArena Spanish | 1393 | 1447 |

## Instruction Following

- GPT-6 Luna: 74.3 (#99)
- Grok 4.20 Multi-Agent: 74.8 (#84)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Instruction Following | 1409 | 1420 |

## Long Context

- GPT-6 Luna: 43.0 (#111)
- Grok 4.20 Multi-Agent: 43.7 (#88)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Longer Query | 1409 | 1431 |

## Writing & Preference

- GPT-6 Luna: 58.3 (#119)
- Grok 4.20 Multi-Agent: 64.0 (#59)

| Benchmark | GPT-6 Luna | Grok 4.20 Multi-Agent |
|---|---|---|
| LMArena Text | 1391 | 1450 |
| LMArena Creative Writing | 1363 | 1436 |
| LMArena Multi-Turn | 1396 | 1452 |

## FAQ

### Is GPT-6 Luna better than Grok 4.20 Multi-Agent?

GPT-6 Luna is the stronger model overall, scoring 53.3 to 46.2 on the Noometry Index.

### Which is cheaper, GPT-6 Luna or Grok 4.20 Multi-Agent?

GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Grok 4.20 Multi-Agent lists at $1.25 and $2.50.

### Is GPT-6 Luna or Grok 4.20 Multi-Agent better for coding?

GPT-6 Luna scores higher on coding benchmarks: 55.5 versus 43.0 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Luna does, with 1.05M tokens against 1M.

### How many benchmarks do GPT-6 Luna and Grok 4.20 Multi-Agent share?

19 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and Grok 4.20 Multi-Agent has 20.